The surface of an experimentally made corium lava flow. Image: from Journeau et al. (2003), Nuclear Engineering and Design. One of the things I enjoy most is running across fascinating information when I'm not even looking for it. Case in point, today's subject. I was doing some research for my class on Fukushima Dai'ichi and Chernobyl when I ran into some references to lava. "Lava?" I thought, "Why are they talking about lava when I thought I was trying to find out about nuclear accidents?" Lo and behold, what do I find but an entire research field that has been making manmade lava for decades. Sure, we've seen some of the recent manmade lava flows done at Syracuse University and small-scale lava in experiments for some time, but here I was finding research that involved a ton (literally) of manmade lava ... and moreover, these lava have been made by accident on a number of occasions with tragic consequences. Let's back up a bit. What I'm talking about here is the result of a meltdown in the core of a nuclear reactor. This is when the nuclear fission reaction occurring within a nuclear reactor is no longer cooled and contained sufficiently to prevent heating of the rods, cases, core containment vessel and anything else nearby, including the concrete floor of the reactor building. When a meltdown begins to occur, as what happened at Chernobyl in 1986 or Fukushima Dai'ichi in 2011, the ability to cool the reactor is insufficient to keep the fuel rods cool, so heat begins to build -- and build rapidly. The two most important primary isotopes used in nuclear fission reactions are uranium-235 and plutonium-239, so it is their fission caused by the absorbtion of a neutron into isotopes with even shorter half-lives (like cesium and strontium) are what produces the heat in the nuclear reactor core. The chain reaction of fission, decays and absorption of the released alpha particles by other atoms is allowed to go unfettered, the heat will build to the point where the fuel rods (made mostly of enriched U, meaning it has more ^235U than the natural distribution of ^235U) will start to bend and, if the heating is allowed to continue, melt. This is usually controlled by cooling water and control rods that can absorb some of neutrons created by fission and decay. However, if there is a problem, the heat can continues to rise and the fuel rods can become fully molten, that is the "meltdown". So, in a sense, a meltdown in a nuclear reactor is the accidental production of lava.
Corium lava flow made experimentally. Image: Argonne National Lab. Now, this lava is, of course, very different than the lava that erupts from a volcano, compositionally. The fuel pellets inside the fuel rods are almost entirely UO2 while the fuel rods in which the pellets are placed is made of zirconium alloys. As the fuel rods heat in an accident, they can get hot enough to start bending (close to 700°C) and if the pellets inside the casing touch, they can begin to melt if the temperature reaches ~1200ºC*. The heat can continue building as the fuel rods melt, eventually forming an entirely molten body that is a mix of the UO2 from the fuel pellets and the zirconium alloy of the casing. If you're going to design safer nuclear reactor, this is where you need to start getting your hands dirty (well, not literally). How does this "corium" (as it is called) behave -- and more importantly, what happens when over components in a reactor come in contact with it? Well, researchers at the Argonne National Lab have created corium in the laboratory in order to see just that (see below). You can check out some great videos of corium lava flowing like pahoehoe (it has an even lower viscosity, which isn't a surprise as it is at 2000ºC, versus 1100-1200ºC for your average basalt) or crusting over when they pour water over it. This lab used upwards of 1 ton** of UO2 lava in some of their experiments to see how quickly corium might melt through the concrete of a nuclear reactor containment vessel (or building). They found that corium lava can melt upwards of 30 cm (12") of concrete in 1 hour! This is why it is so important to know if a nuclear reactor accident has gone into true "meltdown" as the corium lava will rapidly melt its way through the inner containment vessels (or more) in a matter of hours unless it can be cooled again. However, results from these CCI (core-concrete interaction) experiments, suggest that cooling with water may not be sufficient to stop corium from melting the concrete. One thing to remember -- much of the melting of concrete during a meltdown occurs within minutes to hours, so keeping the core cool is vital for stopping the corium for breaching that containment vessel.
Results of an experiment with corium-concrete interactions. Image: Argonne National Lab. Corium lava was produced both during the Chernobyl and Fukushima Dai'ichi accidents (along with minor amounts at Three Mile Island). For the latter, TEPCO, the Japanese energy company who ran Fukushima Dai'ichi, claims that the corium didn't breach the outer wall of the containment vessel (although there is a healthy debate about this). At Chernobyl, there are stunning pictures of corium lavas that melted all the way out of the containment vessel (upwards of 3 meters / 9 feet, see below) -- so these lavas have assimilated concrete and whatever else they could melt on their way out of the containment vessel. This assimilation might actually help in solidifying the corium lava as concrete (which is mostly limestone) has a much lower melting point than corium. Assimilate enough concrete, and the corium should solidify with sufficient cooling -- although research is ongoing about what might be the best composition of concrete for reactors.
Corium lava (solid in the image) that melted through the basement of the Chernobyl nuclear reactor in 1986. So, why is corium so dangerous? Well, even long after the flow has stopped, that lava will be highly radioactive for decades to centuries (along with the surrounding countryside if radioactive material made it out of the containment vessel) as the various radioactive materials in the lava decay. In fact, we don't even have pictures of the corium lava from Fukushima Dai'ichi due to the high levels of radioactivity near the reactor. Instead, measures of radioactivity and gases released from the cooled reactor have been used to model how far the melting of the concrete might have proceeded. In some models, the corium made its way through 0.6 meters (2 feet) of the containment vessel's concrete. Again, cooling the lava by dumping water into the reactor along with assimilation of concrete likely stopped this corium lava flow. Corium is clearly a rare thing -- produced only when humans put a large amount of highly radioactive isotopes together to start of chain reaction. There have been studies that claim that "natural" nuclear reactors (potentially at multiple times) have existed in the Earth's past -- and heck, the dominant source of heat within the Earth comes from the decay of U, thorium and potassium. However, I find it fascinating that manmade lavas have wreaked havoc at least 3 times in the past century as we grapple with how to produce enough energy for the growing demands of the planet. Equally fascinating are the controlled experiments that have tried to come up with ways that we can harness nuclear power more safely, all with these manmade corium lavas. * This is a great example of eutectic melting, where melting begins in places where the two substances touch. The same thing happens when you melt rocks. ** If you do the math, 1 ton of UO2 is actually only about 0.08 m^3 of UO2. Still, I wouldn't want that in my office.